Free to start. Paid when you scale.
The single-node version is not a trial. It is the full product.
Single Node
Free
Forever. Self-hosted.
- ✓Full feature set: graph, lifecycle, MCP, semantic search
- ✓No external database — no Postgres/Qdrant/Redis dependency
- ✓Self-hosted: Docker Compose (bare-metal/binary install coming soon)
- ✓Unlimited memories (single-node hardware limit)
- ✓Community support via GitHub Issues
- ✓License: Functional Source License (FSL)
Multi-Node
Coming soon
On-prem or managed cloud.
- ✓Everything in single-node
- ✓Horizontal sharding across multiple nodes (Business — coming soon)
- ✓Replication and HA failover (Business — coming soon)
- ✓Centralised management backoffice (Business — coming soon)
- ✓Enterprise auth: SSO / SAML (Business — coming soon)
- ✓Audit logs and compliance reporting (Business — coming soon)
- ✓Dedicated support channel
- ✓On-prem via Helm/Kubernetes or managed cloud
- ✓SLA
Frequently asked questions
- Is the free tier really unlimited?
- Yes. Single-node constraints are hardware, not software. Remem does not enforce a memory count limit. How many memories you can store depends on the RAM and disk of the machine you run it on.
- What is the Functional Source License (FSL)?
- FSL is a source-available license that allows unlimited use — including commercial production use — with one restriction: you cannot offer Remem as a competing managed service for others. Every release automatically converts to Apache 2.0 two years after its release date. See fsl.software for the full text.
- Can I use Remem in a commercial product?
- Yes. Internal use is completely unrestricted under FSL. You can build commercial AI products on top of Remem, charge customers for those products, and deploy Remem as part of your infrastructure. The only restriction is operating a managed Remem hosting service in direct competition with Remem.
- What triggers the need for multi-node?
- Single-node scales to millions of memories on modest hardware. Multi-node becomes relevant when you need high availability (zero-downtime failover), horizontal write scaling beyond a single machine's throughput, or centralised multi-tenant management. See the deployment scaling docs for capacity signals.
- GDPR / data residency?
- Because you self-host, all data stays on your infrastructure. Remem makes no outbound network calls (embeddings are generated locally via fastembed). You control where your memories live.
Get notified when Multi-Node ships
No sales call. No qualification. We'll email you when it's ready.